Assess
Each child gets a short one-on-one reading check before sessions begin.
Output: baseline levelReading Corps How we work
Teach at the
right level.
Same grade. Different reading levels. Reading Corps starts by finding what each child can read today, then teaches from there.
600children 5 schools
45 mindaily practice
12 moper cycle
Every child starts at one clear reading level.
Level ladder- Storyfluent reading
- Paragraphconnected text
- Wordword reading
- Letterletters and sounds
- Beginnerstarting point
Progress means moving up the ladder. Results are measured independently and published every cycle.
The Reading Corps model
Five moves. One rule: teach to the child's current level.
The model is simple on purpose. Measure first, group by actual reading level, practice daily, coach tutors, then verify and publish the results.
The two non-negotiables: children are grouped by level, not grade, and results are checked before they are reported.
Group
Children work at Beginner, Letter, Word, Paragraph, or Story level.
Rule: level before gradePractice
Trained local tutors run 45 minutes of structured practice every school day.
Routine: daily and repeatableCoach
A mentor-coordinator observes, supports tutors, and regroups children at midpoint.
Guardrail: quality does not driftVerify
Children are reassessed. A random sample is independently re-tested.
Result: public report + spending5reading levels, from Beginner to Story
45minutes of daily practice
20%minimum random verification sample
40%level-movement threshold before expansion
Why this approach
Not invented here. Borrowed from what works.
Reading Corps adapts Teaching at the Right Level: assess children, group by level, teach daily, coach tutors, and measure again.
+0.7 SD
Large learning gains in randomized trials
Level-based grouping plus daily targeted practice is one of the strongest education findings in the global evidence base.
~$10
Low cost per child in comparable models
The World Bank-backed evidence panel rates level-based instruction with structured guides as highly cost-effective.
+ETH
Already tested inside Ethiopia
TaRL has Ethiopia implementation history, and Oromia has experience with structured Afaan Oromo instruction and EGRA-style assessment.
Evidence trail: J-PAL case study TaRL Africa Ethiopia Pratham TaRL model
Why reading is the gateway
Reading is the hinge. Everything else depends on it.
This section is intentionally simple: five numbers, five reasons early reading matters.
4
Dropout risk rises when children cannot read by Grade 3.
Reading is the skill every later subject assumes.
56%
Ethiopia's primary completion rate leaves little room for children to fall behind early.
Early remediation is cheaper than late recovery.
9-12%
Each additional year of schooling increases earnings.
Reading fluency helps children reach those additional years.
50%
Literate mothers are more likely to send their children to school.
Reading gains compound across generations.
$7-12
Foundational education returns value over a lifetime.
The earlier the intervention, the larger the window.
Source notes
Sources used across this page include UNESCO education and literacy reports, World Bank education and learning poverty publications, ASER Ethiopia, RTI EGRA studies, Psacharopoulos & Patrinos on returns to education, UNICEF child wellbeing reporting, and J-PAL / Pratham / TaRL Africa materials.
What it costs to sponsor one child
One child. Twelve months. Fully tracked. Results published.
$25/month covers one child's full Reading Corps year.
$25 /month
The reading cycle
Twelve months per cycle - with a defined start, midpoint check, and measured close.
A fixed cycle makes the work trackable: baseline, daily practice, midpoint regrouping, endline, public report.
Month 1
Baseline assessment & grouping
Every child assessed. Groups formed by level.
Months 2-5
Daily tutoring begins
45 minutes a day, five days a week.
Month 6
Midpoint check & regroup
Children move groups when ready.
Months 7-11
Tutoring continues
Practice continues at each child's current level.
Month 12
Endline, verify, publish
Reassess, verify a sample, publish results and spending.
Month 1baseline and level groups
Month 6midpoint regrouping
Month 12endline and public report
Theory of change
What has to happen for the model to work.
The pilot is a testable chain, not a promise. Each step reaches the next one only if a condition holds. If a link breaks, we name it and fix it before expanding.
Follow the line down: from today’s classroom to a system that runs level-based reading on its own.
1
Problem
Children at many levels, one curriculum
Grade-level teaching misses children who are not yet reading at grade level.
2
Inputs
$100,000 + bureau partnership + TaRL
Funding, school access, trained tutors, and a tested instructional routine.
3
The engine
Activities
Assess → Group → Tutor daily → Mentor → Verify
The same five moves in every school. Daily, all year.
4
Outcome
≥40%
of children advance one reading level
Bureaus see checked results. Reading movement and actual spending published together.
5
The vision
Long-term effect
Bureau-run level-based instruction, system-wide
The education system adopts what donors helped prove.
What the chain depends on
Each “only if” above is an assumption we track. Six in all. If one slips, we say so.
01
Evidence fit
Does TaRL transfer well here?
02
Tutor supply
Can qualified tutors stay through the cycle?
03
Bureau access
Do agreements hold during school operations?
04
Attendance
Are children present often enough to benefit?
05
Adoption path
Are bureau partners reviewing the evidence?
06 · Stop rule
If results miss the threshold, we do not expand.
No automatic continuation. Every next step is re-earned with checked results.
Accountability
Four commitments - active before the first session, not after.
Child safeguarding
Vetted adults. Open classrooms. Guardian consent for any child image. Aggregate reporting only.
Governance
Board oversight, second sign-off on larger payments, and public filings linked as they are available.
How money moves
No transfer before the legal pathway is finalized. Reports show USD, birr, and the exchange rate used.
Results - what the pilot can and cannot prove
The first cycle shows growth and cost. It does not overclaim causality. Limits are published with the results.
501(c)(3) public charity
EIN 99-3210631
Candid / GuideStar verified
Volunteer-led pilot
100% clarity model